Temporal Awareness, the ability to reason dynamically based on the timestamp when a question is raised, is the key distinction between offline and online video LLMs. Unlike offline models, which rely on complete videos for static, post hoc analysis, online models process video streams incrementally and dynamically adapt their responses based on the timestamp at which the question is posed. Despite its significance, temporal awareness has not been adequately evaluated in existing benchmarks. To fill this gap, we present OVO-Bench (Online-VideO-Benchmark), a novel video benchmark that emphasizes the importance of timestamps for advanced online video understanding capability benchmarking. OVO-Bench evaluates the ability of video LLMs to reason and respond to events occurring at specific timestamps under three distinct scenarios: (1) Backward tracing: trace back to past events to answer the question. (2) Real-time understanding: understand and respond to events as they unfold at the current timestamp. (3) Forward active responding: delay the response until sufficient future information becomes available to answer the question accurately. OVO-Bench comprises 12 tasks, featuring 644 unique videos and approximately human-curated 2,800 fine-grained meta-annotations with precise timestamps. We combine automated generation pipelines with human curation. With these high-quality samples, we further developed an evaluation pipeline to systematically query video LLMs along the video timeline. Evaluations of nine Video-LLMs reveal that, despite advancements on traditional benchmarks, current models struggle with online video understanding, showing a significant gap compared to human agents. We hope OVO-Bench will drive progress in video LLMs and inspire future research in online video reasoning. Our benchmark and code can be accessed at https://github.com/JoeLeelyf/OVO-Bench.

Monotone learning refers to learning processes in which expected performance consistently improves as more training data is introduced. Non-monotone behavior of machine learning has been the topic of a series of recent works, with various proposals that ensure monotonicity by applying transformations or wrappers on learning algorithms. In this work, from a different perspective, we tackle the topic of monotone learning within the framework of Probably Approximately Correct (PAC) learning theory. Following the mechanism that estimates sample complexity of a PAC-learnable problem, we derive a performance lower bound for that problem, and prove the monotonicity of that bound as the sample sizes increase. By calculating the lower bound distribution, we are able to prove that given a PAC-learnable problem with a hypothesis space that is either of finite size or of finite VC dimension, any learning algorithm based on Empirical Risk Minimization (ERM) is monotone if training samples are independent and identically distributed (i.i.d.). We further carry out an experiment on two concrete machine learning problems, one of which has a finite hypothesis set, and the other of finite VC dimension, and compared the experimental data for the empirical risk distributions with the estimated theoretical bound. The results of the comparison have confirmed the monotonicity of learning for the two PAC-learnable problems.

This paper explores and assesses in what ways generative AI can assist in translating survey instruments. Writing effective survey questions is a challenging and complex task, made even more difficult for surveys that will be translated and deployed in multiple linguistic and cultural settings. Translation errors can be detrimental, with known errors rendering data unusable for its intended purpose and undetected errors leading to incorrect conclusions. A growing number of institutions face this problem as surveys deployed by private and academic organizations globalize, and the success of their current efforts depends heavily on researchers' and translators' expertise and the amount of time each party has to contribute to the task. Thus, multilinguistic and multicultural surveys produced by teams with limited expertise, budgets, or time are at significant risk for translation-based errors in their data. We implement a zero-shot prompt experiment using ChatGPT to explore generative AI's ability to identify features of questions that might be difficult to translate to a linguistic audience other than the source language. We find that ChatGPT can provide meaningful feedback on translation issues, including common source survey language, inconsistent conceptualization, sensitivity and formality issues, and nonexistent concepts. In addition, we provide detailed information on the practicality of the approach, including accessing the necessary software, associated costs, and computational run times. Lastly, based on our findings, we propose avenues for future research that integrate AI into survey translation practices.

Retrieval-augmented generation (RAG) is a powerful technique that enhances downstream task execution by retrieving additional information, such as knowledge, skills, and tools from external sources. Graph, by its intrinsic "nodes connected by edges" nature, encodes massive heterogeneous and relational information, making it a golden resource for RAG in tremendous real-world applications. As a result, we have recently witnessed increasing attention on equipping RAG with Graph, i.e., GraphRAG. However, unlike conventional RAG, where the retriever, generator, and external data sources can be uniformly designed in the neural-embedding space, the uniqueness of graph-structured data, such as diverse-formatted and domain-specific relational knowledge, poses unique and significant challenges when designing GraphRAG for different domains. Given the broad applicability, the associated design challenges, and the recent surge in GraphRAG, a systematic and up-to-date survey of its key concepts and techniques is urgently desired. Following this motivation, we present a comprehensive and up-to-date survey on GraphRAG. Our survey first proposes a holistic GraphRAG framework by defining its key components, including query processor, retriever, organizer, generator, and data source. Furthermore, recognizing that graphs in different domains exhibit distinct relational patterns and require dedicated designs, we review GraphRAG techniques uniquely tailored to each domain. Finally, we discuss research challenges and brainstorm directions to inspire cross-disciplinary opportunities.

Reinforcement learning (RL) in the real world necessitates the development of procedures that enable agents to explore without causing harm to themselves or others. The most successful solutions to the problem of safe RL leverage offline data to learn a safe-set, enabling safe online exploration. However, this approach to safe-learning is often constrained by the demonstrations that are available for learning. In this paper we investigate the influence of the quantity and quality of data used to train the initial safe learning problem offline on the ability to learn safe-RL policies online. Specifically, we focus on tasks with spatially extended goal states where we have few or no demonstrations available. Classically this problem is addressed either by using hand-designed controllers to generate data or by collecting user-generated demonstrations. However, these methods are often expensive and do not scale to more complex tasks and environments. To address this limitation we propose an unsupervised RL-based offline data collection procedure, to learn complex and scalable policies without the need for hand-designed controllers or user demonstrations. Our research demonstrates the significance of providing sufficient demonstrations for agents to learn optimal safe-RL policies online, and as a result, we propose optimistic forgetting, a novel online safe-RL approach that is practical for scenarios with limited data. Further, our unsupervised data collection approach highlights the need to balance diversity and optimality for safe online exploration.

This paper introduces ActPC-Geom, an approach to accelerate Active Predictive Coding (ActPC) in neural networks by integrating information geometry, specifically using Wasserstein-metric-based methods for measure-dependent gradient flows. We propose replacing KL-divergence in ActPC's predictive error assessment with the Wasserstein metric, suggesting this may enhance network robustness. To make this computationally feasible, we present strategies including: (1) neural approximators for inverse measure-dependent Laplacians, (2) approximate kernel PCA embeddings for low-rank approximations feeding into these approximators, and (3) compositional hypervector embeddings derived from kPCA outputs, with algebra optimized for fuzzy FCA lattices learned through neural architectures analyzing network states. This results in an ActPC architecture capable of real-time online learning and integrating continuous (e.g., transformer-like or Hopfield-net-like) and discrete symbolic ActPC networks, including frameworks like OpenCog Hyperon or ActPC-Chem for algorithmic chemistry evolution. Shared probabilistic, concept-lattice, and hypervector models enable symbolic-subsymbolic integration. Key features include (1) compositional reasoning via hypervector embeddings in transformer-like architectures for tasks like commonsense reasoning, and (2) Hopfield-net dynamics enabling associative long-term memory and attractor-driven cognitive features. We outline how ActPC-Geom combines few-shot learning with online weight updates, enabling deliberative thinking and seamless symbolic-subsymbolic reasoning. Ideas from Galois connections are explored for efficient hybrid ActPC/ActPC-Chem processing. Finally, we propose a specialized HPC design optimized for real-time focused attention and deliberative reasoning tailored to ActPC-Geom's demands.

Large language models (LLMs) have made great progress in classification and text generation tasks. However, they are mainly trained on English data and often struggle with low-resource languages. In this study, we explore adding a new language, i.e., Persian, to Llama (a model with a limited understanding of Persian) using parameter-efficient fine-tuning. We employ a multi-stage approach involving pretraining on monolingual Persian data, aligning representations through bilingual pretraining and instruction datasets, and instruction-tuning with task-specific datasets. We evaluate the model's performance at each stage on generation and classification tasks. Our findings suggest that incorporating the Persian language, through bilingual data alignment, can enhance classification accuracy for Persian tasks, with no adverse impact and sometimes even improvements on English tasks. Additionally, the results highlight the model's initial strength as a critical factor when working with limited training data, with cross-lingual alignment offering minimal benefits for the low-resource language. Knowledge transfer from English to Persian has a marginal effect, primarily benefiting simple classification tasks.

Domain-adaptive post-training of large language models (LLMs) has emerged as a promising approach for specialized domains such as medicine and finance. However, significant challenges remain in identifying optimal adaptation criteria and training strategies across varying data and model configurations. To address these challenges, we introduce FINDAP, a systematic and fine-grained investigation into domain-adaptive post-training of LLMs for the finance domain. Our approach begins by identifying the core capabilities required for the target domain and designing a comprehensive evaluation suite aligned with these needs. We then analyze the effectiveness of key post-training stages, including continual pretraining, instruction tuning, and preference alignment. Building on these insights, we propose an effective training recipe centered on a novel preference data distillation method, which leverages process signals from a generative reward model. The resulting model, Llama-Fin, achieves state-of-the-art performance across a wide range of financial tasks. Our analysis also highlights how each post-training stage contributes to distinct capabilities, uncovering specific challenges and effective solutions, providing valuable insights for domain adaptation of LLMs. Project page: https://github.com/SalesforceAIResearch/FinDap

Large vision-language models (LVLMs) augment language models with visual understanding, enabling multimodal reasoning. However, due to the modality gap between textual and visual data, they often face significant challenges, such as over-reliance on text priors, hallucinations, and limited capacity for complex visual reasoning. Existing benchmarks to evaluate visual reasoning in LVLMs often rely on schematic or synthetic images and on imprecise machine-generated explanations. To bridge the modality gap, we present DrivingVQA, a new benchmark derived from driving theory tests to evaluate visual chain-of-thought reasoning in complex real-world scenarios. It offers 3,931 expert-crafted multiple-choice problems and interleaved explanations grounded with entities relevant to the reasoning process. We leverage this dataset to perform an extensive study of LVLMs' ability to reason about complex visual scenarios. Our experiments reveal that open-source and proprietary LVLMs struggle with visual chain-of-thought reasoning under zero-shot settings. We investigate training strategies that leverage relevant entities to improve visual reasoning. Notably, we observe a performance boost of up to 7\% when reasoning over image tokens of cropped regions tied to these entities.

Reinforcement learning has become an essential algorithm for generating complex robotic behaviors. However, to learn such behaviors, it is necessary to design a reward function that describes the task, which often consists of multiple objectives that needs to be balanced. This tuning process is known as reward engineering and typically involves extensive trial-and-error. In this paper, to avoid this trial-and-error process, we propose the concept of Constraints as Rewards (CaR). CaR formulates the task objective using multiple constraint functions instead of a reward function and solves a reinforcement learning problem with constraints using the Lagrangian-method. By adopting this approach, different objectives are automatically balanced, because Lagrange multipliers serves as the weights among the objectives. In addition, we will demonstrate that constraints, expressed as inequalities, provide an intuitive interpretation of the optimization target designed for the task. We apply the proposed method to the standing-up motion generation task of a six-wheeled-telescopic-legged robot and demonstrate that the proposed method successfully acquires the target behavior, even though it is challenging to learn with manually designed reward functions.